Searches for Point-Like Sources of Neutrinos with the 40-String Icecube Detector
Total Page:16
File Type:pdf, Size:1020Kb
Searches for Point-like Sources of Neutrinos with the 40-String IceCube Detector by Jonathan P. Dumm A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Physics) at the University of Wisconsin – Madison 2011 c 2011 Jonathan P. Dumm All Rights Reserved Searches for Point-like Sources of Neutrinos with the 40-String IceCube Detector Jonathan P. Dumm Under the supervision of Professor Teresa Montaruli At the University of Wisconsin – Madison The IceCube Neutrino Observatory is the first 1 km3 neutrino telescope. Data were collected using the partially-completed IceCube detector in the 40-string config- uration recorded between 2008 April 5 and 2009 May 20, totaling 375.5 days livetime. An unbinned maximum likelihood ratio method is used to search for astrophysical sig- nals. The data sample contains 36,900 events: 14,121 from the northern sky, mostly muons induced by atmospheric neutrinos and 22,779 from the southern sky, mostly high energy atmospheric muons. The analysis includes time-integrated searches for individual point sources and targeted searches for specific stacked source classes and spatially extended sources. While this analysis is sensitive to TeV–PeV energy neu- trinos in the northern sky, it is primarily sensitive to neutrinos with energy greater than about 1 PeV in the southern sky. A number of searches are performed and sig- nificances (given as p-values, the chance probability to occur with only background present) calculated: (1) a scan of the entire sky for point sources (p=18%), (2) a predefined list of 39 interesting source candidates (p=62%), (3) stacking 16 sources of TeV gamma rays observed by Milagro and Fermi, along with an unconfirmed hot spot (p=32%), (4) stacking 127 starburst galaxies (p=100%), and (5) stacking five nearby galaxy clusters (p=78%). No evidence for a signal is found in any of the searches. Limits are set for neutrino fluxes from astrophysical sources over the entire sky and compared to predictions. The sensitivity is at least a factor of two better than previ- ous searches (depending on declination), with 90% confidence level muon neutrino flux upper limits being between E2dN/dE 2 200 10−12 TeV cm−2 s−1 in the northern ∼ − × sky and between 3 700 10−12 TeV cm−2 s−1 in the southern sky. The stacked source − × searches provide the best limits to specific source classes. For the case of supernova remnants, we are just a factor of three from ruling out realistic predictions. The full IceCube detector is expected to improve the sensitivity to E−2 sources by another factor of two in the first year of operation. Teresa Montaruli (Adviser) i Abstract The IceCube Neutrino Observatory is the first 1 km3 neutrino telescope. Data were collected using the partially-completed IceCube detector in the 40-string config- uration recorded between 2008 April 5 and 2009 May 20, totaling 375.5 days livetime. An unbinned maximum likelihood ratio method is used to search for astrophysical sig- nals. The data sample contains 36,900 events: 14,121 from the northern sky, mostly muons induced by atmospheric neutrinos and 22,779 from the southern sky, mostly high energy atmospheric muons. The analysis includes time-integrated searches for individual point sources and targeted searches for specific stacked source classes and spatially extended sources. While this analysis is sensitive to TeV–PeV energy neu- trinos in the northern sky, it is primarily sensitive to neutrinos with energy greater than about 1 PeV in the southern sky. No evidence for a signal is found in any of the searches. A number of searches are performed and significances (given as p-values, the chance probability to occur with only background present) calculated: (1) a scan of the entire sky for point sources (p=18%), (2) a predefined list of 39 interesting source candidates (p=62%), (3) stacking 16 sources of TeV gamma rays observed by Milagro and Fermi, along with an unconfirmed hot spot (p=32%), (4) stacking 127 starburst galaxies (p=100%), and (5) stacking five nearby galaxy clusters, testing four different models for the CR distribution (p=78%). Limits are set for neutrino fluxes from astrophysical sources over the entire sky and compared to predictions. The sensitivity is at least a factor of two better than previous searches (depending on declination), with 90% confidence level muon neutrino flux upper limits being be- tween E2dN/dE 2 200 10−12 TeV cm−2 s−1 in the northern sky and between ∼ − × ii 3 700 10−12 TeV cm−2 s−1 in the southern sky. The stacked source searches provide − × the best limits to specific source classes. For the case of supernova remnants, we are just a factor of three from ruling out realistic predictions. The full IceCube detector is expected to improve the sensitivity to E−2 sources by another factor of two in the first year of operation. iii Acknowledgements I dedicate this work to my brother, Dennis. He taught me so much, and I’m still trying to figure out how much of it is actually true. I would first like to thank my parents, Brad and Sue, for working so hard to provide endless opportunities for me. I would also like to thank my wife, Jess, for her love and support through all this work. I am little without her. I owe a debt of gratitude to my advisor, Teresa Montaruli, not only because of her guidance and support for six years, but also because of her dedication to the field of neutrino astronomy. She is truly inspiring. Thanks to my early mentors, Juande Zornoza and, in particular, Chad Finley, for introducing me to the world of likelihood analysis and for his immense intellectual curiosity. Also David Boersma, whose expertise in computing and generosity with time have helped so many. It has been a pleasure working closely with Mike Baker, Juanan Aguilar, and Naoko Kurahashi on the point source analysis. I would like to thank Hagar Landsman for introducing me to DOM testing and preparing me to work on hardware at the South Pole. Thanks to all the others who make Madison such a terrific place to work on IceCube and a terrific place to live. Finally, it takes many people to make IceCube a reality. I would like to thank iv all members of the IceCube Collaboration for their dedication over the years. v Contents Abstract i Acknowledgements iii 1 Neutrino Astronomy and the High Energy Universe 1 1.1 TheNeutrino ................................ 2 1.2 SolarNeutrinos ............................... 3 1.3 Supernova1987A .............................. 5 1.4 CosmicRays................................. 6 1.4.1 Cosmic Ray Flux and Composition . 6 1.4.2 Cosmic Ray Acceleration . 9 1.4.3 Candidate Cosmic Ray Accelerators . 11 1.5 AstrophysicalNeutrinoProduction . 14 1.6 DiffuseNeutrinoAstronomyResults. 16 2 Principles of Neutrino Detection 20 2.1 Neutrino-nucleon Interactions . 20 2.2 OtherNeutrinoInteractions . 22 vi 2.3 ChargedLeptonPropagation. 23 2.3.1 Electrons .............................. 23 2.3.2 Muons................................ 24 2.3.3 TauParticles ............................ 25 2.3.4 CerenkovRadiation.ˇ . 26 2.4 CosmicRayBackgrounds. 27 2.5 TheEarthasaNeutrinoTarget . 30 2.6 NeutrinoOscillations ............................ 32 3 The IceCube Neutrino Observatory 36 3.1 DigitalOpticalModules . 37 3.2 DataAcquisition .............................. 40 3.3 Calibration ................................. 43 3.4 Installation ................................. 44 3.5 OpticalPropertiesofSouthPoleIce. 44 3.5.1 HoleIce ............................... 46 3.6 OtherNeutrinoTelescopes . 47 3.6.1 AMANDA.............................. 47 3.6.2 Deep Sea Telescopes . 48 4 Event Reconstruction and Selection 50 4.1 HitPreparation ............................... 50 4.2 TrackReconstruction............................ 51 4.2.1 Line-FitFirstGuessReconstruction. 52 4.2.2 Likelihood Reconstruction . 53 vii 4.2.2.1 Pandel Function . 56 4.3 EventSelection ............................... 58 4.3.1 FilteringLevels ........................... 59 4.3.1.1 Online (Level 1) Filter . 59 4.3.1.2 Offline (Level 2) Filter . 61 4.3.1.3 Final Event Selection . 68 5 Simulation and Comparison to Data 74 5.1 EventGeneration .............................. 75 5.1.1 NeutrinoSimulation . 75 5.1.2 Atmospheric Muon Simulation . 76 5.2 Propagation................................. 77 5.2.1 Chargedleptons........................... 77 5.2.2 CerenkovPhotons..........................ˇ 77 5.3 DetectorSimulation............................. 77 5.4 DataandMCComparisons . 78 5.5 AtmosphericNeutrinoSpectrum. 81 5.6 DetectorPerformance. 82 6 Search Method 87 6.1 MaximumLikelihoodMethod . 87 6.1.1 SignalPDF ............................. 88 6.1.2 BackgroundPDF .......................... 90 6.1.3 TestStatistic ............................ 91 6.2 HypothesisTesting ............................. 91 viii 6.3 Calculating Significance and Discovery Potential . ....... 92 6.4 Calculating Upper Limits and Sensitivity . 93 6.4.1 Including Systematic Errors in Upper Limits . 99 6.5 MeasuringSpectralIndexandCutoffSpectra . 100 6.6 ModificationforStackingSources . 101 6.7 ModificationforExtendedSources . 103 7 Searches for Neutrino Sources 106 7.1 All-skyScan.................................107 7.2 APrioriSourceList ............................107 7.3 MilagroTeVSourceStacking . 107 7.4 StarburstGalaxyStacking . 108 7.5 GalaxyClusterStacking . .109 8 Systematic Errors 112 9 Results 115 9.1 All-skyScan.................................115 9.2 APrioriSourceList ............................117 9.3 StackingSearches..............................120 10 Implications for